KR20080004012A - An video encoding/decoding method and apparatus - Google Patents

Abstract

A video encoding method and apparatus, and a video decoding method and apparatus are provided to perform prediction encoding by using correlation between plural color components composing a video, thereby improving the encoding efficiency of a video. A video encoding method comprises the following steps of: performing prediction encoding of a pixel block of a predetermined size of a first color component video out of two color components in an input video(1010); performing motion prediction of a remaining color component pixel block corresponding to the first color component video to generate a first prediction pixel block candidate of the remaining color component pixel block(1020); restoring the first prediction-encoded color component pixel block(1030); predicting the remaining color component pixel block by using the restored remaining color component pixel block and generating a second prediction pixel block candidate; and determining one of the first prediction pixel block candidate and the second prediction pixel block candidate as the prediction pixel block of the remaining color component pixel block.

Description

An video encoding / decoding method and apparatus

1A to 1C are diagrams illustrating an example of an R (Red) color component image, a G (Green) color component image, and a B (Blue) color component image constituting one color image.

FIG. 2A is a graph illustrating a correlation between the G color component image and the B color component image of FIGS. 1B and 1C.

3 is a block diagram showing the configuration of a video encoding apparatus according to a preferred embodiment of the present invention.

4 is a block diagram illustrating in detail the configuration of the prediction pixel block generator 314 of FIG. 3.

5 is a flowchart illustrating an image encoding method according to an exemplary embodiment of the present invention.

6A and 6B are diagrams for describing a process of generating a first prediction pixel block candidate of the remaining color component pixel blocks by the first prediction pixel block candidate generator 314a of FIG. 4.

FIG. 7A shows a 16 × 16 pixel block of a G color component image included in an input image, FIG. 7B shows a 16 × 16 pixel block of a B color component image included in the input image, and FIG. 7C shows an R color component image provided in the input image. Shows a 16 by 16 pixel block.

8A is a diagram illustrating a processing sequence of 8 × 8 pixel blocks in the image encoding method and apparatus according to the present invention, and FIG. 8B is a diagram illustrating a processing sequence of 4 × 4 pixel blocks in the video encoding method and apparatus according to the present invention. to be.

9 is a diagram illustrating a configuration of an image decoding apparatus according to a preferred embodiment of the present invention.

10 is a flowchart illustrating an image decoding method according to a preferred embodiment of the present invention.

The present invention relates to encoding and decoding of an image. More particularly, the present invention relates to motion prediction information of one color component image or a reconstructed one by using correlation between a plurality of color components constituting the image. An image encoding method and apparatus, and a decoding method and apparatus for improving encoding efficiency by predicting another color component image from a color component image.

In general, when acquiring an image, the first image is acquired in the form of an RGB color format. When encoding an RGB color format image, it is generally converted to a color format such as YUV (or YCbCr). In this case, Y is a black-and-white image and luminance component, and U (or Cb) and V (or Cr) have color components. In the RGB image, the information is distributed evenly in R, G, and B. In the YUV (or YCbCr) image, the information is concentrated in the Y component, and the amount of information is reduced in the U (or Cb) and V (or Cr). Therefore, the compression has the advantage that the compression efficiency is increased. In order to further improve the compression efficiency, a YUV (or YCbCr) 4: 2 sampled by sampling the chromatic components U (or Cb) and V (or Cr) of the YUV (or YCbCr) image in 1/4 size is generally used. : 0 Use video.

However, sampling U (or Cb) and V (or Cr) in 1/4 sizes in a YUV (or YCbCr) 4: 2: 0 image produces color distortion and is not suitable for high quality applications. Therefore, there is a need for a method of effectively encoding a YUV (or YCbCr) 4: 4: 4 image without U (or Cb) and V (or Cr) sampling. Recently, a technique called residual color transform (RCT), which directly encodes an RGB 4: 4: 4 image, has been proposed to remove color distortion generated when converting an RGB image into YUV (or YCbCr). have.

As described above, in case of directly encoding an image having the same resolution between color components such as YUV (or YCbCr) 4: 4: 4 and RGB 4: 4: 4 image, the coding efficiency is lowered when the conventional encoding method is applied. Therefore, when encoding YUV (or YCbCr) 4: 4: 4 image or encoding RGB input image without conversion into YUV (or YCbCr) as it is in the RGB domain, the prediction is performed according to the statistical characteristics of the image to maintain high image quality. There is a need for a method for improving coding efficiency.

SUMMARY OF THE INVENTION The present invention has been made in view of the correlation between each color component of RGB without converting an RGB color format input image to another color format, and the motion prediction information of one of the color components of RGB or the restored one color An object of the present invention is to provide an encoding method and apparatus for decoding an image, and a decoding method and apparatus for improving encoding efficiency by predicting an image of another color component from the component image.

In order to solve the above technical problem, an image encoding method according to the present invention includes (a) prediction encoding of a pixel block having a predetermined size of a first color component image among images of at least two color components included in an input image. Performing; (b) generating a first prediction pixel block candidate of the remaining color component pixel blocks by performing motion prediction on the remaining color component pixel blocks corresponding to the first color component pixel block; (c) reconstructing the predictively coded first color component pixel block; (d) generating a second prediction pixel block candidate by predicting the remaining color component pixel blocks using the reconstructed first color component pixel blocks; And (e) determining one of the first prediction pixel block candidate and the second prediction pixel block candidate as a prediction pixel block of the remaining color component pixel blocks.

According to an embodiment of the present invention, an image encoding apparatus predicts a motion of a pixel block of a remaining color component corresponding to a pixel block of a predetermined size of a first color component image previously predicted and encoded among images of at least two color components included in an input image. A first prediction pixel block candidate generator configured to generate a first prediction pixel block candidate of the remaining color component pixel blocks by performing a? A reconstruction unit for reconstructing the predictively coded first color component pixel block; A second prediction pixel block candidate generator configured to predict the remaining color component pixel blocks using the reconstructed first color component pixel blocks to generate a second prediction pixel block candidate of the remaining color component pixel blocks; And a determination unit configured to determine one of the first prediction pixel block candidate and the second prediction pixel block candidate as a prediction pixel block candidate of the remaining color component pixel blocks.

An image decoding method according to the present invention includes: (a) decoding a pixel block having a predetermined size of a first color component image among encoded images of at least two color components included in an input bitstream; (b) the remainder through motion prediction for the remaining color component pixel blocks corresponding to the first color component pixel block or prediction using the decoded first color component pixel block according to the prediction mode information included in the bitstream. Generating a prediction pixel block of the color component pixel block; And (c) decoding the remaining color component pixel blocks using the generated prediction pixel blocks of the remaining color component pixel blocks.

An apparatus for decoding an image according to the present invention includes: a first decoder configured to decode a pixel block having a predetermined size of a first color component image among encoded images of at least two color components included in an input bitstream; The remaining color component pixels through motion prediction or prediction using the decoded first color component pixel block corresponding to the first color component pixel block according to the prediction mode information included in the bitstream. A prediction pixel block generation unit generating a prediction pixel block of the block; And a second decoder which decodes the pixel blocks of the remaining color component images by using the generated prediction pixel blocks of the remaining color component pixel blocks.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1A to 1C are diagrams illustrating examples of an R (Red) color component image, a G (Green) color component image, and a B (Blue) color component image constituting one color image, respectively, and FIGS. 2A are FIGS. 1C is a graph illustrating a correlation between a G color component image and a B color component image, and FIG. 2B is a graph showing a correlation between the R color component image and the G color component image of FIGS. 1A and 1B.

In general, when encoding a color image, prediction encoding is performed for each color component image to remove overlapping information in each color component. Referring to FIGS. 1A to 1C, pixel values of RGB color component images of the same position constituting one color image have similar pixel values, which can be reconfirmed through the correlation graphs illustrated in FIGS. 2A and 2B. have. In addition, a strong correlation exists between each pixel block of the RGB color component image after performing motion prediction and compensation.

Therefore, according to the present invention, a predetermined first color component image of a plurality of color component images constituting the image is subjected to predictive encoding according to a general image encoding scheme, for example, the H.264 standard, and then each color component image. The motion vector generated from the motion prediction result of the first color component image is considered in consideration of the correlation therebetween to perform motion prediction and compensation of the other remaining color component images, or the other remaining color component images from the reconstructed first color component image. Provided are an encoding method and an apparatus for predicting. As an example, if an image includes three color components of RGB, the image encoding method and apparatus according to the present invention first predictively encode a G color component image, and then determine a motion vector determined at the time of motion estimation of the G color component image. Motion estimation and compensation of the corresponding R color component image and B color component image are performed using the same motion vector, or the R color component image and B color component image are predicted or reconstructed using the reconstructed G color component image. The R color component image is predicted using the G color component image, and the B color component image is predicted using the reconstructed R color component image. In the above-described example, the coding order of the color component image is not limited to the above example and may be changed as much as possible.

According to another embodiment, the image encoding method and apparatus according to the present invention first predictively encode a G color component image, and then perform motion prediction and compensation on the R color component image and the B color component image, or reconstruct the G color component. The R color component image and the B color component image are predicted using the image. Alternatively, first, the G color component image is predictively encoded, and then motion prediction and compensation are performed on the R color component image and the B color component image, or the R color component image is predicted and reconstructed using the reconstructed G color component image. The B color component image is predicted using the R color component image. In the above-described example, the coding order of the color component image is not limited to the above example and may be changed as much as possible.

3 is a block diagram showing the configuration of a video encoding apparatus according to a preferred embodiment of the present invention. Hereinafter, for the convenience of description, a description will be given of an image encoding apparatus based on the H.264 standard. However, the image encoding apparatus according to the present invention may be applied to other image encoding apparatuses that perform residue coding.

Referring to FIG. 3, the image encoding apparatus 300 may include a motion estimator 302, a motion compensator 304, an intra predictor 306, a subtractor 307, a transformer 308, and a quantizer 309. ), Reordering unit 310, entropy coding unit 311, inverse quantization unit 312, inverse transform unit 313, prediction pixel block generation unit 314, adder 315, filter 316, frame memory 317 and the control unit 318 are provided.

The motion estimator 302 and the motion compensator 304 perform inter prediction for searching the prediction pixel block of the current pixel block in a previous or subsequent reference frame. The motion estimator 302 transmits the motion vector determined as a result of the motion prediction for the pixel block of the first color component image to the motion compensator 304 and the prediction pixel block generator 314. As described later, the motion vector of the first color component pixel block is used as a motion vector for motion prediction and compensation of the remaining color component pixel blocks corresponding to the first color component pixel block. As described above, it is also possible to generate the motion vector through the motion prediction of the remaining color component pixel blocks without using the motion vector of the first color component pixel block.

The intra predictor 306 performs intra prediction to predict the predicted pixel block of the current pixel block from the current frame. In detail, the intra predictor 306 divides the first color component image into pixel blocks having a predetermined size. The intra prediction unit 306 may use the intra prediction modes available for the divided first color component pixel blocks according to their sizes, for example, the intra 16 × 16 prediction mode, the intra 4 × 4 prediction mode, and the intra 8 × 8. Intra prediction is performed according to the prediction mode.

The subtractor 307 generates a first residue by subtracting the prediction pixel block generated through inter prediction or intra prediction from the corresponding pixel block of the input first color component original image. The generated first residue is converted into a frequency domain by the converter 308 and quantized by the quantizer 309. The transform coefficients of the quantized first residue component are rearranged by the reordering unit 310, and then encoded by the entropy coding unit 314 and output in a bitstream form.

The transformed and quantized first residue is inversely quantized and inversely transformed by the inverse quantizer 312 and the inverse transformer 313. The adder 315 reconstructs the first color component pixel block by adding the first quantized and inverse transformed first residue and the first color component prediction pixel block. The reconstructed first color component image is passed through a filter 316 to perform deblocking filtering, and then stored in the frame memory 317 and used to perform inter prediction on the next frame. In addition, the reconstructed first color component pixel block is input to the intra predictor 306 and used as a reference value for intra prediction of the next pixel block. Also, the reconstructed first color component pixel block is input to the prediction pixel block generator 314 to predict the remaining color component pixel blocks.

4 is a block diagram illustrating in detail the configuration of the prediction pixel block generator 314 of FIG. 3.

The prediction pixel block generator 314 uses the motion vector generated as a result of the motion prediction of the first color component image by using the correlation between the color component images constituting the color image to perform motion prediction and compensation of other remaining color component images. By generating a first prediction pixel block candidate of the remaining color component pixel blocks, generating a second prediction pixel block candidate of the remaining color component pixel blocks from the reconstructed first color component image, and generating the generated first prediction pixel block candidate. And compares the costs of the second prediction pixel block candidates to determine and output a final prediction pixel block of the remaining color component pixel blocks.

Referring to FIG. 4, the prediction pixel block generator 314 includes a first prediction pixel block candidate generator 314a, a second prediction pixel block candidate generator 314b, and a determiner 314c.

The first prediction pixel block candidate generator 314a predicts the motion of the remaining color component pixel blocks corresponding to the first color component pixel block by using the motion vector of the first color component pixel block input from the motion estimation unit 302. And performing compensation to generate a first prediction pixel block candidate of the remaining color component pixel blocks. As described above, the first prediction pixel block candidate of the remaining color component pixel blocks may be generated by performing motion prediction and compensation of the remaining color component pixel blocks.

The second prediction pixel block candidate generator 314b predicts a corresponding color component pixel block from the pixel block of a predetermined size of the reconstructed first color component image. 2A and 2B, the pixel values of the color component images constituting the color image have a correlation with each other and can be linearly modeled as a linear function. The second prediction pixel block candidate generator 314b uses a predictor generated through a linear modeling process to restore the first color component by using the restored pixel value of the first color component pixel block as a parameter. A second prediction pixel block candidate of the remaining color component pixel blocks corresponding to the pixel block is generated. Details of the linear modeling process for predicting the remaining color component pixel blocks from the pixel blocks of the reconstructed first color component image will be described later.

When the second prediction pixel block candidate generator 314b predictively encodes an input image including three or more color components, such as an RGB color input image, the second prediction pixel block candidate generator 314b reconstructs all remaining second and third color component pixel blocks. A second prediction pixel block candidate of each of the second and third color component pixel blocks may be generated by predicting using the pixel block of the one color component image, or the second prediction pixel block candidate of the second color component pixel block is A prediction may be generated from the reconstructed first color component pixel block and a second prediction pixel block candidate of the third color component pixel block may be generated from the reconstructed second color component pixel block. That is, the second prediction pixel block candidate generator 314b generates a second prediction pixel block candidate of the remaining color component pixel blocks from the reconstructed first color component pixel block, or the remaining color component pixel blocks previously processed and reconstructed. May generate second prediction pixel block candidates of the other remaining color component pixel blocks from.

The determiner 314c determines the final prediction pixel block of the remaining color component pixel blocks by comparing the costs of the first prediction pixel block candidate and the second prediction pixel block candidate of the remaining color component pixel blocks. For example, the determiner 314c may apply a well-known rate-distortion cost to calculate the distortion of the bit amount and the image quality for each of the first and second prediction pixel block candidates, and then calculate the two values. The prediction pixel block candidate having the smallest value is added as the final prediction pixel block. In this case, the bit amount may be calculated by performing entropy encoding after converting and quantizing the bit amount. The distortion of the image quality may be the average of the sum of squared difference values from the original image by reconstructing the image. As another embodiment, the final prediction pixel block may be selected by comparing only the distortion cost in order to reduce the amount of computation.

Referring to FIG. 3 again, the subtractor 307 subtracts the second and third color component prediction pixel blocks generated by the prediction pixel block generator 314 from the pixel blocks of the second and third color component original images. Generate a second and third residue. Like the first residue described above, the second and third residues are encoded through a transform, quantization, and entropy encoding process, and are output in a bitstream form.

In addition, the transformed and quantized second and third residues are inversely quantized and inversely transformed by the inverse quantizer 312 and the inverse transformer 313, and the adder 315 is inversely quantized and inversely transformed. The third residue and the second and third color component prediction pixel blocks generated by the prediction pixel block generation unit 314 are added to reconstruct the second and third color component pixel blocks. The reconstructed second and third color component images are passed through a filter 316 that performs deblocking filtering, and are then stored in the frame memory 317 and used to perform inter prediction on the next frame. When the third color component pixel block is predicted from the pixel block of the reconstructed second color component image, the second color component pixel block reconstructed by the adder 315 is input to the prediction pixel block generator 314 again. .

The controller 318 controls each component of the image encoding apparatus 300 and determines a prediction mode of the current pixel block. In detail, the control unit 318 calculates the cost of the predicted image using the correlation between the inter predicted image, the intra predicted image, and the color component image according to the present invention, and has the smallest cost among the predicted images. The prediction mode is determined as the final prediction mode. When the cost of the predicted image is greater than a predetermined threshold, the controller 318 encodes each color component image through inter prediction or intra prediction according to the prior art instead of the predictive encoding method according to the present invention. You can choose.

In addition, the controller 318 calculates a reference value indicating a correlation between each color component image as shown in FIGS. 2A and 2B, and when the reference value between each color component image is equal to or less than a predetermined threshold, prediction according to the present invention. Instead of the encoding method, each color component image may be selected to be encoded through inter prediction or intra prediction according to the prior art. Here, as a reference value, dispersion or standard deviation indicating a scatter diagram between each color component may be used.

5 is a flowchart illustrating an image encoding method according to an exemplary embodiment of the present invention. Hereinafter, an operation and an image encoding method of an image encoding apparatus according to the present invention will be described with reference to FIGS. 3 to 5.

In operation 510, the intra predictor 306 performs prediction encoding on a pixel block of a predetermined size of a first color component image selected from among input images having a plurality of input color components. As described above, the motion estimator 302 and the motion compensator 304 perform motion prediction and compensation on the first color component pixel block to generate a temporally predicted pixel block of the first color component pixel block. The predictor 306 generates a spatial predicted pixel block by performing intra prediction on the first color component pixel block. The control unit 318 calculates the distortion of the bit amount and the image quality of the temporal prediction pixel block and the spatial prediction pixel block of the first color component pixel block, and finally predicts the prediction pixel block having the smallest cost of the first color component pixel block. Determined by the pixel block. The first residue, which is the difference between the determined final prediction pixel block and the first color component original pixel block, is output in a bitstream form through transformation, quantization, and entropy encoding.

In operation 520, the first prediction pixel block candidate generator 314a of the prediction pixel block generator 314 may convert the motion vector of the first color component pixel block input from the motion estimation unit 302 into the first color component pixel block. The pixel block of the reference frame indicated by the determined motion vector of the remaining color component pixel block is determined as the first prediction pixel block candidate of the remaining color component pixel block.

6A and 6B are diagrams for describing a process of generating a first prediction pixel block candidate of the remaining color component pixel blocks by the first prediction pixel block candidate generator 314a of FIG. 4. In FIG. 6A, reference numeral 610 denotes a G color component pixel block coded in a G color component current frame G _t , and reference numeral 611 refers to a G color component referred to when a motion of the G color component pixel block 610 is predicted. A G color component pixel block at the same position as the G color component pixel block 610 of the current frame in the frame G _{t -1} is indicated by reference numeral 612 to denote a G color component reference pixel block determined as a result of the motion prediction. In FIG. 6B, reference numeral 620 denotes an R color component pixel block coded in an R color component current frame R _t , and reference numeral 621 denotes an R color component referred to when a motion of the R color component pixel block 610 is predicted. An R color component pixel block at the same position as the R color component pixel block 620 of the current frame in the reference frame R _{t −1} is indicated by reference numeral 622 to denote an R color component reference pixel block determined as a result of motion prediction.

If the first color component is called G color, the first prediction pixel block candidate generator 314a may have the same motion as the motion vector MV _G of the G color component pixel block 610 generated by the motion estimation unit 302. The vector is determined as the motion vector MV _R of the corresponding R color component pixel block 620. In addition, the first prediction pixel block candidate generator 314a may have an R color component that is equal in time to the G color component reference frame G _{t −1} referred to by the motion vector MV _G of the G color component pixel block 610. In the reference frame R _{t -1} , the current R blocks the reference pixel block 622 pointed to by the motion vector MV _R of the same R color component pixel block as the motion vector MV _G of the G color component pixel block 610. The first prediction pixel block candidate of the color component pixel block is determined. Similarly, the first prediction pixel block candidate generator 314a may have the same B color in time as the G color component reference frame G _{t −1} referred to by the motion vector MV _G of the G color component pixel block 610. The first prediction of the current B color component pixel block indicates a reference pixel block indicated by the motion vector MV _B of the B color component pixel block that is the same as the motion vector MV _G of the G color component pixel block 610 in the component reference frame. Determined as a pixel block candidate. In this case, motion vectors corresponding to the B and R color component pixel blocks need not be transmitted to the decoding end. Meanwhile, as described above, the motion estimation unit 302 may determine the pixel block candidate of the first prediction pixel block corresponding to the motion vector by performing motion prediction of the B and R color component pixel blocks. Motion vector information corresponding to the B and R color component pixel blocks should be transmitted to the decoding end.

In operation 530, the first residue generated during the prediction encoding on the first color component pixel block is inversely quantized and inversely transformed, and the first color component predicted pixel block is inversely quantized and inversely transformed to add the first residue. Restore the one color component pixel block.

In operation 540, the second prediction pixel block candidate generator 314b generates a second prediction pixel block of the remaining color component pixel blocks by using the reconstructed first color component pixel block.

FIG. 7A shows a 16 × 16 pixel block of a G color component image included in an input image, FIG. 7B shows a 16 × 16 pixel block of a B color component image included in an input image, and FIG. 7C shows an R color component image provided in an input image. Shows a 16 by 16 pixel block. Here, g _{i, j} , b _{i, j} , r _{i, j} denote pixel values located in the i th row and the j th column of the 16 × 16 pixel block of the G, B, and R color component images, respectively. In addition, the pixels hatched in each of FIGS. 7A to 7C represent reconstructed pixels of the neighboring pixel block processed before the current pixel block.

, 상기

에 대응되는 B 색 성분 화소 블록의 i번째 행, j번째 열에 위치한 화소의 예측값을

, a는 소정의 가중치, b는 소정의 오프셋 값이라 할 때, 제 2 예측 화소 블록 후보 생성부(314b)는 복원된 G 색 성분 화소 블록의 화소값과 대응되는 B 색 화소 블록의 화소값의 상관 관계를 다음의 수학식 1과 같이 1차 함수로 모델링하고, G 색 성분 화소 블록의 복원된 화소값을 매개 변수로 하여 대응되는 B 색 성분 화소 블록의 화소값을 예측한다.The pixel values located in the i th row and j th column of the restored G color component 16 × 16 pixel block.

, remind

Predicted values of pixels located in the i th row and the j th column of the B-color component pixel block corresponding to

When a is a predetermined weight and b is a predetermined offset value, the second prediction pixel block candidate generator 314b determines the pixel value of the B color pixel block corresponding to the pixel value of the reconstructed G color component pixel block. The correlation is modeled using a linear function as shown in Equation 1 below, and the pixel value of the corresponding B color component pixel block is predicted using the restored pixel value of the G color component pixel block as a parameter.

The prediction pixel values obtained through Equation 1 are clipped to an integer value between 0 and 255 when each pixel value of the image is represented by 8 bits. Although the values of a and b may be changed according to the position (i, j) of the pixel, in the embodiment of the present invention, the case of having a constant value within a predetermined block is considered.

According to an embodiment, the values of a and b are predetermined functions using the restored pixel values of the pixel block around the G color component and the restored pixel values of the pixel block around the B color component as shown in Equations 2 and 3 below. It is determined by the value.

A and b may be defined in various forms. As an example, a and b are prediction pixel values of the B color component peripheral pixel block predicted through Equation 1 using the reconstructed pixel values of the pixel block surrounding the G color component, and the restored pixels of the pixel block surrounding the B color component. Is determined so that the difference with the values is minimal. As another example, a may be defined as 1 and b may be defined as an average of the difference between the restored pixel values of the pixel block around the B color component and the restored pixel values of the pixel block around the G color component as shown in Equation 4 below. have.

In the above-described determination of the values of a and b, pixel values of neighboring pixel blocks adjacent to the current pixel block are used. The pixels of the neighboring pixel block used in the determination of the values of a and b may be employed more or less than pixels adjacent to the upper or left side of the current pixel block. At this time, the values of a and b need not be transmitted to the decoding end. This is because the decoding end can generate the same a and b in the same manner as the same coding end.

Meanwhile, as another embodiment, the values of a and b may be determined using pixel values in the current pixel block without using neighboring pixel blocks. For example, as shown in Equation 5, the values of a and b are predicted through the original pixel values b _{i, j} and 16 in the pixel block of 16 × 16 size of the B color component image. A sum of diff (r _B ) of a difference between prediction values of a corresponding B color component image may be determined to be a value that is minimized.

Here, p is an integer of 1 or more.

As another example, the value of a is determined to be 1, and b is the original pixel values b _{i, j} and the reconstructed G color component 16 × 16 in the B color component 16 × 16 pixel block, as shown in Equation 6 below. The average of the difference between the pixel values G ′ _{i, j} of the pixel block may be determined.

In such embodiments, it is necessary to transmit the values of a and b to the decoder.

As described above, when the values of the constants a and b of Equation 1 are determined, the second prediction pixel block candidate generator 314b determines each pixel value _{g'i, j} of the reconstructed G color component pixel block. By substituting Equation 1, a second prediction pixel block candidate is generated by predicting pixel values of a corresponding B-color component pixel block.

Another method for determining the values of a and b may be a method based on a linear regression model, which is widely used in statistics.

8A is a diagram illustrating a processing sequence of 8 × 8 pixel blocks in the image encoding method and apparatus according to the present invention, and FIG. 8B is a diagram illustrating a processing sequence of 4 × 4 pixel blocks in the video encoding method and apparatus according to the present invention. to be.

The second prediction pixel block candidate generator 314b may generate the second prediction pixel block candidate of the remaining color component pixel blocks in 8 × 8 block mode and 4 × 4 block mode in addition to the 16 × 16 block mode described above. .

Referring to FIG. 8A, when a pixel block of a B color component image is processed in an 8 × 8 mode, 8 × 8 pixel blocks of four B color component images are sequentially predicted from left to right and top to bottom. . The processing of the 8 × 8 pixel block of the B color component image has only changed the block size, and similarly to the process of estimating the pixel value of the 16 × 16 pixel block of the B color component image, the B color component is represented by Equation 1. Each pixel value in an 8x8 pixel block of an image may be predicted.

That is, when processing in the 8x8 mode, a and b are the predicted pixel values and the B color of the pixel block around the B color component predicted through Equation 1 using the reconstructed pixel values of the pixel block around the G color component. It is determined as a value that minimizes the difference between the reconstructed pixel values of the component peripheral pixel block, or a is 1, b is the reconstructed pixel values of the pixel block of the color component B and G as shown in Equation 7 below. It may be defined as an average of the differences between the restored pixel values of the pixel block around the color component.

Also, a and b are sums of differences between prediction values of the B color component image corresponding to the original pixel values b _{i, j} in the 8 × 8 pixel block of the B color component image (Sum of diff (r _B )). ) Is determined to be minimized, or a is 1, and b is the original pixel values b _{i, j} and 8 reconstructed G color components in the 8 × 8 pixel block of the B color component image as shown in Equation 8 below. It may be determined as an average of the differences between the pixel values G ′ _{i, j} of the × 8 pixel block.

Referring to FIG. 8B, when a pixel block of a B color component image is processed in a 4 × 4 mode, 16 B color component 4 × 4 pixel blocks are sequentially predicted from left to right and top to bottom. The pixel values of each 4x4 pixel block of the B color component image differ only in block size, and are similar to the process of predicting pixel values of the 16x16 pixel block or 8x8 pixel block of the B color component image described above. Can be predicted through Equation 1.

In other words. a and b are the predicted pixel values of the pixel block around the B color component predicted by Equation 1 using the restored pixel values of the pixel block around the G color component, and the restored pixel values of the pixel block around the B color component. Is determined to be the minimum value, or a is 1 and b is the restored pixel values of the pixel block around the B color component and the restored pixel values of the pixel block around the G color component as shown in Equation 9 below. It can be defined as the average of the differences between.

In addition, a and b of Equation 1 are sums of differences of prediction values of the B color component image corresponding to the original pixel values b _{i, j} in the 4 × 4 pixel block of the B color component image. _B )) is determined as a value to be minimized, or a is fixed to 1 and b is reconstructed from the original pixel values b _{i, j} in the 4x4 pixel block of the B color component image as shown in Equation 10 below. The average value of the difference between the pixel values G ′ _{i, j} of the 4 × 4 pixel block of the G color component image may be determined.

As described above, the second prediction pixel block of the B color component pixel block may be configured in units of 16 × 16, 8 × 8, and 4 × 4 blocks. As an adaptive embodiment, correlation prediction of each macroblock may be performed in units of one suitable block among the three block modes.

The subtractor 307 generates a second residue by calculating a difference between the predicted pixel blocks predicted by the second predicted pixel block generator 314b from the pixel blocks of the B color component original image, and generates the generated second residue. Is output in the form of a bitstream through transform, quantization, and entropy encoding.

Next, the pixel values of the pixel block of the R color component image may be predicted using the pixel values of the pixel block of the G color component image reconstructed similarly to the pixel values of the pixel block of the B color component image.

On the other hand, when the second prediction pixel block candidate generator 314b predicts the pixel value of the pixel block of the R color component image, the pixel block of the B color component image previously processed instead of the pixel block of the reconstructed G color component image. The second prediction pixel block of the R color component pixel block may be generated using the reconstructed pixel value of. That is, the inverse transform and inverse quantization of the value obtained by transforming and quantizing the second residue, which is the difference between the predicted pixel blocks of the B color component image, from the pixel block of the B color component original image, By adding the prediction pixel blocks of the B color component image, the reconstructed pixel blocks of the B color component can be used to predict the pixel blocks of the R color component image.

, 상기

에 대응되는 R 색 성분 영상의 화소 블록의 i번째 행, j번째 열에 위치한 화소의 예측값을

, c는 B 색 성분 영상과 R 색 성분 사이의 상관 관계를 나타내는 소정의 가중치, d는 소정의 오프셋 값이라 할 때, 제 2 예측 화소 블록 후보 생성부(314b)는 복원된 B 색 성분 화소 블록의 화소값과 R 색 성분 화소 블록 사이의 상관 관계를 다음의 수학식 11과 같이 1차 함수로 모델링하여, B 색 성분 영상의 화소값에 대응되는 R 색 성분 영상의 화소값을 예측한다.Specifically, the pixel values located in the i th row and the j th column of the pixel block of the restored B color component image are determined.

, remind

Predicted values of pixels located in the i th row and the j th column of the pixel block of the R color component image corresponding to

, c is a predetermined weight indicating a correlation between a B color component image and an R color component, and d is a predetermined offset value, so that the second prediction pixel block candidate generator 314b reconstructs the restored B color component pixel block. The correlation between the pixel value and the R color component pixel block is modeled using a linear function as shown in Equation 11 to predict the pixel value of the R color component image corresponding to the pixel value of the B color component image.

The prediction pixel values obtained through Equation 11 are clipped to an integer value between 0 and 255 when each pixel value of the image is represented by 8 bits. The values of c and d may be determined similarly to the values of a and b.

When the values of the constants c and d of Equation 11 are determined, the second prediction pixel block candidate generator 314b converts each pixel value b ′ _{i, j} of the reconstructed B color component pixel block to Equation 11; By substituting and predicting pixel values of the corresponding R color component pixel block, a second prediction pixel block candidate of the R color component pixel is generated.

9 is a diagram illustrating a configuration of an image decoding apparatus according to a preferred embodiment of the present invention.

9, the image decoding apparatus 900 according to the present invention includes an entropy decoder 910, a reordering unit 920, an inverse quantization unit 930, an inverse transform unit 940, an adder 950, and intra prediction. A unit 960, a motion compensator 970, a predicted pixel block generator 980, and a filter 990 are provided.

The entropy decoder 910 and the reordering unit 920 receive the compressed bitstream and perform entropy decoding to generate quantized coefficients. The inverse quantization unit 930 and the inverse transformer 940 perform inverse quantization and inverse transformation on the quantized coefficients to extract residue information, motion vector information, and prediction mode information of each color component image. Here, the prediction mode information may include predetermined syntax information indicating whether the bit stream is encoded according to the intra prediction method according to the present invention. In the case of a bitstream encoded according to the image encoding method according to the present invention, the prediction mode information may include predictor information used to predict pixel values of pixel blocks of the remaining color component images.

When the current pixel block is an intra predicted pixel block, the intra predictor 960 generates a prediction block by using the previously decoded neighboring pixel block. The motion compensator 970 generates a prediction block through motion compensation when the current pixel block is an inter predicted pixel block.

When the prediction pixel block generator 980 is a bitstream encoded according to the above-described image encoding method of the present invention, the prediction pixel block generation unit 980 uses the motion vector of the decoded first color component pixel block to move the corresponding color component pixel block. Prediction pixel blocks of the remaining color component pixel blocks are generated by performing prediction or prediction using the decoded first color component pixel blocks.

Specifically, similarly to the prediction pixel block generator 314 of FIG. 3, the prediction pixel block generator 980 of FIG. 9 is the same as the motion vector of the first color component pixel block according to the prediction mode information included in the bitstream. A vector is determined as a motion vector of the remaining color component pixel blocks, and a pixel block of a reference frame indicated by the determined motion vector of the remaining color component pixel blocks is determined as a prediction pixel block. In addition, the prediction pixel block generation unit 980 estimates pixel values of the remaining color component pixel blocks by substituting the pixel values of the first color component pixel blocks reconstructed according to the prediction mode information included in the bitstream into Equation 1. A prediction pixel block of the remaining color component pixel blocks may be generated. In this case, when a and b are predicted using neighboring pixel blocks, a and b do not need to be included in the bitstream. That is, in this case, a and b are not included in the prediction mode information. If the third color component image is predicted using the reconstructed second color component image, the pixel values of the pixel block of the reconstructed second color component image are substituted into Equation 11 to predict the pixel of the third color component pixel block. You can create a block.

The adder 950 generates the prediction pixel block and the inverse transformer 940 generated by any one of the intra predictor 960, the motion compensator 970, and the predictive pixel block generator 980 according to the prediction mode of the current pixel block. Each of the color component pixel blocks is decoded by summing the residues of the respective color component pixel blocks.

10 is a flowchart illustrating an image decoding method according to a preferred embodiment of the present invention.

Referring to FIG. 10, in operation 1010, a bitstream including encoded images of at least two color components is received, and decoding of a first color component pixel block is performed among images of a plurality of color components included in the bitstream. Perform. Specifically, the pixel block of the first color component image is decoded by adding the predicted pixel block of the inter or intra predicted first color component pixel block and the decoded first residue.

In operation 1020, the pixel values of the remaining color component pixel blocks are predicted using the decoded first color component pixel blocks and the prediction mode information included in the header of the bitstream to generate the prediction pixel blocks of the remaining color component pixel blocks. As described above, the prediction pixel block generation unit 980 generates a predictor as shown in Equation 1 through the prediction mode information, substitutes pixel values of the decoded first color component pixel block, and displays the remaining second and third colors. The pixel values of the pixel block of the component image are predicted, or a predictor such as Equation 11 is generated and the pixel values of the decoded second color component pixel block are substituted to predict the pixel values of the third color component pixel block. In addition, according to the prediction mode, it is also possible to generate a prediction pixel block by motion compensation of the second and third color component pixel blocks.

In operation 1030, the second and third color component pixel blocks are decoded by adding each of the prediction pixel blocks of the second and third color component pixel blocks to each of the inversely transformed second and third residue components.

The invention can also be embodied as computer readable code on a computer readable recording medium. The computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like, and may also be implemented in the form of a carrier wave (for example, transmission over the Internet). Include. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

So far I looked at the center of the preferred embodiment for the present invention. Those skilled in the art will understand that the present invention may be implemented in a modified form without departing from the essential characteristics of the present invention. Therefore, the disclosed embodiments should be considered in descriptive sense only and not for purposes of limitation. The scope of the present invention is shown in the claims rather than the foregoing description, and all differences within the scope will be construed as being included in the present invention.

According to the present invention described above, the encoding efficiency of an image can be improved by performing predictive encoding by using a correlation between a plurality of color component images forming one image.

In addition, according to the present invention, by encoding in the RGB domain as it is without converting the RGB input image to the YUV domain, by preventing the distortion of the color generated during the process of converting the RGB input image to another color format, the image Can improve the quality.

Claims (52)

In the video encoding method, (a) performing predictive encoding on a pixel block of a predetermined size of a first color component image among images of at least two color components included in the input image; (b) generating a first prediction pixel block candidate of the remaining color component pixel blocks by performing motion prediction on the remaining color component pixel blocks corresponding to the first color component pixel block; (c) reconstructing the predictively coded first color component pixel block; (d) generating a second prediction pixel block candidate by predicting the remaining color component pixel blocks using the reconstructed first color component pixel blocks; And (e) determining one of the first prediction pixel block candidate and the second prediction pixel block candidate as a prediction pixel block of the remaining color component pixel blocks. The method of claim 1 wherein the color components are An encoding method characterized by being an R (Red) color component, a G (Green) color component, and a B (Blue) color component. The method of claim 1, wherein step (a) (a1) generating a first color component prediction pixel block by performing motion prediction and / or intra prediction on the first color component pixel block; (a2) generating a first residue which is a difference between the generated first color component prediction pixel block and a first color component original pixel block; And (a2) further comprising transforming, quantizing and entropy encoding the first residue. The method of claim 1, wherein step (b) (b1) determining a motion vector identical to a motion vector of the first color component pixel block generated as a result of the motion prediction performed during the prediction encoding of the first color component pixel block, as the motion vector of the remaining color component pixel blocks; And and (b2) determining the pixel block of the reference frame indicated by the determined motion vector as the first prediction pixel block candidate of the remaining color component pixel blocks. The method of claim 1, wherein step (c) (c1) inverse quantization and inverse transformation of a first residue, which is a difference between the transformed and quantized first color component prediction pixel block and the first color component original pixel block; And and (c2) reconstructing the first color component pixel block by adding the first quantized and inverse transformed first residue and the first color component prediction pixel block. The method of claim 1, wherein step (d) The size of the restored first color component pixel block is i × j (i, j is an integer), and the restored pixel value located in the i th row and the j th column of the restored first color component pixel block.

, remind

The prediction pixel value of the second prediction pixel block candidate of the second color component pixel block corresponding to

When a is a predetermined weight and b is a predetermined offset value, the following equation;

And a second prediction pixel block candidate of the second color component pixel block is generated by using a method. The method of claim 6, wherein a and b are The difference between the predicted pixel values of the second color component peripheral pixel block predicted using the restored pixel values of the first color component peripheral pixel block and the restored pixel values of the second color component peripheral pixel block is minimal. The encoding method characterized in that the value to be. The method of claim 6, A is 1, and b is an average of a difference between the reconstructed pixel values of the pixel block around the second color component and the reconstructed pixel values of the pixel block around the first color component. . The method of claim 6, wherein a and b are Original pixel values of the second color component pixel block (

) And prediction pixel values of the second prediction pixel block candidate of the second color component pixel block

And a value that minimizes the sum of the absolute difference between the values. The method of claim 6, A is 1, and b is the original pixel values of the second color component pixel block (

) And the reconstructed pixel values of the first color component pixel block (

Coding method characterized in that the average of the difference between the (). The method of claim 6, wherein a and b are A coding method, characterized in that it is determined based on a linear regression model. The method of claim 6, wherein step (d) Generate a second prediction pixel block candidate of the second color component pixel block using the reconstructed first color component pixel block, and generate a second prediction pixel block candidate of the third color component pixel block using the reconstructed second color component pixel block. A coding method characterized by generating a prediction pixel block candidate. The method of claim 12, wherein step (d) The size of the restored second color component pixel block is i × j (i, j is an integer), and the pixel values located in the i th row and the j th column of the restored second color component pixel block.

, remind

The prediction pixel value of the second prediction pixel block candidate of the third color component pixel block corresponding to

When c is a predetermined weight and d is a predetermined offset value, the following equation;

And a second prediction pixel block candidate of the third color component pixel block is generated by using a method. The method of claim 13, wherein c and d are The difference between the predicted pixel values of the third color component peripheral pixel block predicted using the restored pixel values of the second color component peripheral pixel block and the restored pixel values of the third color component peripheral pixel block is different. A coding method, characterized in that the minimum value. The method of claim 13, C is 1, and d is an average of a difference between the restored pixel values of the pixel block surrounding the third color component and the restored pixel values of the pixel block surrounding the second color component. . The method of claim 13, wherein c and d are The original pixel value of the third color component pixel block

) And prediction pixel values of the second prediction pixel block candidate of the third color component pixel block

And a value that minimizes the sum of the absolute difference between the values. The method of claim 13, C is 1, and d is the original pixel value of the third color component pixel block (

) And the reconstructed pixel value of the second color component pixel block

Coding method characterized in that the average of the difference between the). In the video encoding apparatus, The remaining color component pixels are performed by performing motion prediction on the remaining color component pixel blocks corresponding to pixel blocks having a predetermined size among the images of at least two color components included in the input image. A first prediction pixel block candidate generator generating a first prediction pixel block candidate of the block; A reconstruction unit for reconstructing the predictively coded first color component pixel block; A second prediction pixel block candidate generator configured to predict the remaining color component pixel blocks using the reconstructed first color component pixel blocks to generate a second prediction pixel block candidate of the remaining color component pixel blocks; And a determining unit configured to determine one of the first prediction pixel block candidate and the second prediction pixel block candidate as a prediction pixel block candidate of the remaining color component pixel blocks. 19. The method of claim 18, wherein the color components are An encoding device characterized by being an R (Red) color component, a G (Green) color component, and a B (Blue) color component. 19. The method of claim 18, wherein the first prediction pixel block candidate generator is The same motion vector as the motion vector of the first color component pixel block generated as a result of the motion prediction performed during the prediction encoding of the first color component pixel block is determined as the motion vector of the remaining color component pixel blocks, and the determined motion vector is determined. And determine a pixel block of a reference frame indicated by the first prediction pixel block candidate of the remaining color component pixel blocks. 19. The method of claim 18, wherein the second prediction pixel block candidate generator The size of the restored first color component pixel block is i × j (i, j is an integer), and the restored pixel value located in the i th row and the j th column of the restored first color component pixel block.

, remind

The prediction pixel value of the second prediction pixel block candidate of the second color component pixel block corresponding to

When a is a predetermined weight and b is a predetermined offset value, the following equation;

And a second prediction pixel block candidate of the second color component pixel block is generated by using a quantum symbol. The method of claim 21, wherein a and b are The difference between the predicted pixel values of the second color component peripheral pixel block predicted using the restored pixel values of the first color component peripheral pixel block and the restored pixel values of the second color component peripheral pixel block is minimal. The encoding device characterized in that the value to be. The method of claim 21, A is 1, and b is an average of a difference between the reconstructed pixel values of the pixel block around the second color component and the reconstructed pixel values of the pixel block around the first color component. . The method of claim 21, wherein a and b are Original pixel values of the second color component pixel block (

) And prediction pixel values of the second prediction pixel block candidate of the second color component pixel block

And a value that minimizes the sum of the absolute difference between the values. The method of claim 21, A is 1, and b is the original pixel values of the second color component pixel block (

) And the reconstructed pixel values of the first color component pixel block (

Encoding device, characterized in that the average of the difference between the (). The method of claim 21, wherein a and b are A coding apparatus, characterized in that determined based on a linear regression model. The method of claim 21, wherein the second prediction pixel block candidate generator The second prediction pixel block candidate of the second color component pixel block is generated using the reconstructed first color component pixel block, and the second prediction of the third color component pixel block is generated using the reconstructed second color component pixel block. An encoding device characterized by generating a pixel block candidate. 28. The method of claim 27, wherein the second prediction pixel block candidate generator is The size of the restored second color component pixel block is i × j (i, j is an integer), and the pixel values located in the i th row and the j th column of the restored second color component pixel block.

, remind

The prediction pixel value of the second prediction pixel block candidate of the third color component pixel block corresponding to

When c is a predetermined weight and d is a predetermined offset value, the following equation;

And a second prediction pixel block candidate of the third color component pixel block is generated. 29. The method of claim 28, wherein c and d are The difference between the predicted pixel values of the third color component peripheral pixel block predicted using the restored pixel values of the second color component peripheral pixel block and the restored pixel values of the third color component peripheral pixel block is different. The encoding device characterized in that the minimum value. The method of claim 28, C is 1, and d is an average of a difference between the restored pixel values of the pixel block surrounding the third color component and the restored pixel values of the pixel block surrounding the second color component. . 29. The method of claim 28, wherein c and d are The original pixel value of the third color component pixel block

) And prediction pixel values of the second prediction pixel block candidate of the third color component pixel block

And a value that minimizes the sum of the absolute difference between the values. The method of claim 28, C is 1, and d is the original pixel value of the third color component pixel block (

) And the reconstructed pixel value of the second color component pixel block

Encoding device, characterized in that the average of the difference between the). In the video decoding method, (a) decoding a pixel block having a predetermined size of a first color component image among encoded images of at least two color components included in the input bitstream; (b) the remainder through motion compensation or prediction using the decoded first color component pixel block corresponding to the first color component pixel block according to the prediction mode information included in the bitstream. Generating a prediction pixel block of the color component pixel block; And (c) decoding the remaining color component pixel blocks by using the predicted pixel blocks of the generated remaining color component pixel blocks. 34. The method of claim 33, wherein the color components are A decoding method comprising R (Red) color component, G (Green) color component, and B (Blue) color component. The method of claim 33, wherein step (a) (a1) decoding a first residue of the first color component pixel block included in the bitstream; (a2) generating a prediction pixel block of the first color component pixel block by performing intra prediction or motion compensation according to the prediction mode of the first color component pixel block; And and (a3) decoding the first color component pixel block by adding the decoded first residue and a prediction pixel block of the first color component pixel block. The method of claim 33, wherein step (b) When the motion compensation for the remaining color component pixel blocks is performed, the same vector as the motion vector of the first color component pixel block is determined as the motion vector of the remaining color component pixel blocks, and the pixel block of the reference frame indicated by the determined motion vector is determined. The decoding method, characterized in that for determining the prediction pixel block. The method of claim 33, wherein step (b) The size of the decoded first color component pixel block is i × j (i, j is an integer), and the value of the decoded pixel located at the i th row and the j th column of the decoded first color component pixel block.

, remind

The prediction pixel value of the prediction pixel block of the second color component pixel block corresponding to

When a is a predetermined weight and b is a predetermined offset value, the following equation;

And generating a predicted pixel block of the second color component pixel block by using a quantum symbol. 38. The method of claim 37, wherein a and b are And inserting the bitstream into the bitstream when the video is encoded. The method of claim 33, wherein step (b) A prediction pixel block of a second color component pixel block is generated using the decoded first color component pixel block with respect to the second color component pixel block among the remaining color component pixel blocks, and a first one of the remaining color component pixel blocks is generated. For the three color component pixel block, a decoding pixel block of the third color component pixel block is generated using the decoded second color component pixel block. The method of claim 39, wherein step (b) I × j (i, j is an integer) of the size of the decoded second color component pixel block, and pixel values located in an i th row and a j th column of the decoded second color component pixel block.

, remind

The prediction pixel value of the prediction pixel block of the third color component pixel block corresponding to

When c is a predetermined weight and d is a predetermined offset value, the following equation;

And generating a predicted pixel block of the third color component pixel block by using a spectral method. 41. The method of claim 40, wherein c and d are And inserting the bitstream into the bitstream when the video is encoded. The method of claim 33, wherein step (c) And decoding the remaining color component pixel blocks by adding residues of the remaining color component pixel blocks and the prediction pixel blocks of the remaining color component pixel blocks included in the bitstream. In the video decoding apparatus, A first decoder configured to decode a pixel block having a predetermined size of a first color component image among encoded images of at least two color components included in the input bitstream; The remaining color component pixels through motion compensation for the remaining color component pixel blocks corresponding to the first color component pixel block or prediction using the decoded first color component pixel blocks according to the prediction mode information included in the bitstream. A prediction pixel block generation unit generating a prediction pixel block of the block; And And a second decoder configured to decode the pixel block of the remaining color component image by using the generated prediction pixel block of the remaining color component pixel block. 44. The method of claim 43, wherein the color components are And an R (Red) color component, a G (Green) color component, and a B (Blue) color component. 44. The apparatus of claim 43, wherein the first decoder Decoding a first residue of the first color component pixel block included in the bitstream and performing intra prediction or motion compensation according to a prediction mode of the first color component pixel block And generating a prediction pixel block, and decoding the first color component pixel block by adding the decoded first residue and the prediction pixel block of the first color component pixel block. 45. The method of claim 43, wherein the prediction pixel block generator When the motion compensation for the remaining color component pixel blocks is performed, the same vector as the motion vector of the first color component pixel block is determined as the motion vector of the remaining color component pixel blocks, and the pixel block of the reference frame indicated by the determined motion vector is determined. And determine the predicted pixel block as the prediction pixel block. 45. The method of claim 43, wherein the prediction pixel block generator The size of the decoded first color component pixel block is i × j (i, j is an integer), and the value of the decoded pixel located at the i th row and the j th column of the decoded first color component pixel block.

, remind

The prediction pixel value of the prediction pixel block of the second color component pixel block corresponding to

When a is a predetermined weight and b is a predetermined offset value, the following equation;

And a predictive pixel block of the second color component pixel block is generated by using a quantum symbol. 48. The method of claim 47, wherein a and b are And inserting the bitstream into the bitstream at the time of encoding the video. 45. The method of claim 43, wherein the prediction pixel block generator A prediction pixel block of a second color component pixel block is generated using the decoded first color component pixel block with respect to the second color component pixel block among the remaining color component pixel blocks, and a first one of the remaining color component pixel blocks is generated. A decoding device according to claim 3, wherein a prediction pixel block of the third color component pixel block is generated using the decoded second color component pixel block. 50. The apparatus of claim 49, wherein the prediction pixel block generation unit I × j (i, j is an integer) of the size of the decoded second color component pixel block, and pixel values located in an i th row and a j th column of the decoded second color component pixel block.

, remind

The prediction pixel value of the prediction pixel block of the third color component pixel block corresponding to

When c is a predetermined weight and d is a predetermined offset value, the following equation;

And a predictive pixel block of the third color component pixel block is generated by using a quantum symbol. 51. The method of claim 50, wherein c and d are And inserting the bitstream into the bitstream at the time of encoding the video. The method of claim 33, wherein the second decoding unit And decoding the remaining color component pixel blocks by adding residues of the remaining color component pixel blocks and the prediction pixel blocks of the remaining color component pixel blocks included in the bitstream.

Images